Does the Pistorm600 still let you use PCMCIA port?

[Hiddenevil]

I've been having a devil of a time with my Furia card and been eyeing up the Pistorm as a possible alternative. However I watched a YT video this morning with someone installing a Pistorm and they couldn't get their CF to PCMCIA adapter to work after the install.

Does the Pistorm kill the PCMCIA port?

[alexh]

It shouldn't do. It is emulating a full 32-bit accelerator and so does not need to map it's RAM to the Zorro II space where the PCMCIA is.

If it doesn't work I would suspect it is the PiStorm configuration.

With the Furia I thought you just hold down the keys CTRL+A+A for more than 3 seconds to disable 4MB of RAM during warm reboot

https://www.kuchinka.cz/furia/

[supaduper]

Originally Posted by alexh

It shouldn't do. It is emulating a full 32-bit accelerator and so does not need to map it's RAM to the Zorro II space where the PCMCIA is.

If it doesn't work I would suspect it is the PiStorm configuration.

With the Furia I thought you just hold down the keys CTRL+A+A for more than 3 seconds to disable 4MB of RAM during warm reboot

https://www.kuchinka.cz/furia/

there is some clever software put there that lets you keep the 9mb and run the PCMCIA , works great, info in this thread

https://eab.abime.net/showthread.php?t=117513

[derSammler]

there is some clever software put there that lets you keep the 9mb and run the PCMCIA , works great, info in this thread

The issue is with Kickstart 3.1 mainly, in which Commodore did a stupid mistake. With anything below 3.1, there was no issue with PCMCIA and 8 MB of Fast RAM. But with 3.1, they completely disabled the PCMCIA port with more than 4 MB, even if all that would have been needed was to disable the memory mapping of SRAM cards.

You can make it work (again) by simply using card.res from 3.0.

The PiStorm has no issue with PCMCIA at all.

[Daedalus]

Unfortunately, it's impossible to completely eliminate the risk of an inserted card corrupting memory, so the only sensible approach is to disable the slot when conflicting fast RAM is found. Hence Commodore fixed the problem in 3.1. OS 3.1.4 tried to work around the issue but even that was found to cause problems, so it reverted to the 3.1 solution in 3.2.

[ncafferkey]

Originally Posted by Daedalus

Unfortunately, it's impossible to completely eliminate the risk of an inserted card corrupting memory, so the only sensible approach is to disable the slot when conflicting fast RAM is found. Hence Commodore fixed the problem in 3.1. OS 3.1.4 tried to work around the issue but even that was found to cause problems, so it reverted to the 3.1 solution in 3.2.

Does the problem only occur if the card responds to common memory space accesses? And does it take a driver enabling that region for corruption to occur?

[Daedalus]

The exact details are buried in one of the OS 3.1.4 development threads, but the general issue is that it's something that would happen very early, at the Budgie/Gayle hardware config stage before the OS is even properly running, let alone activating drivers. From what I remember, before being initialised as I/O cards, some cards would indeed respond to accesses in the memory area as if they were RAM cards.

[ncafferkey]

@Daedalus: thanks.